Yes, a very odd combination of quality and quantity with Fostex and Pyle Pro. As many of you know, the Pyle Pro PPA15 has become very popular as an open baffle sub driver. Decided to give it a shot, but crossed over higher. I chose two pairs instead of the goldwood 18 because of increased surface area, reduced footprint, and higher excursion capability. All for roughly the same price (Paid 130 for both pairs, check amazon for great price!).
The full range driver will be a Fostex FF85KeN, from Dave of course. He was gracious enough to sell me a cosmetically challenged pair to fit my budget as a college student. I was originally going to run them passive, but plans have changed and it will be active instead.
Read over MJK's open baffle papers (thank you!) and decided on a U-Baffle with short wings (4-5 inches) to push the ripple well above the crossover point. The FF85K will be crossed over at 250hz-300hz and the PPA15s will be crossed over at around 150hz to combat rising response. Both Linkwitz-Riley 12db/octave.
The top of the front baffle roughly follows the 2.2* Driver diameter rule that has been floating around here. Its about 6.75 inches wide at the centerline of the driver. The FR will be mounted higher than normal of 32 inches to 38 inches (preference). Here are my preliminary drawings. I don't think I will be changing them much. I apologize for the dimension clutter.
Any thoughts or suggestions would be appreciated. I will update this thread once the units have been built!
The full range driver will be a Fostex FF85KeN, from Dave of course. He was gracious enough to sell me a cosmetically challenged pair to fit my budget as a college student. I was originally going to run them passive, but plans have changed and it will be active instead.
Read over MJK's open baffle papers (thank you!) and decided on a U-Baffle with short wings (4-5 inches) to push the ripple well above the crossover point. The FF85K will be crossed over at 250hz-300hz and the PPA15s will be crossed over at around 150hz to combat rising response. Both Linkwitz-Riley 12db/octave.
The top of the front baffle roughly follows the 2.2* Driver diameter rule that has been floating around here. Its about 6.75 inches wide at the centerline of the driver. The FR will be mounted higher than normal of 32 inches to 38 inches (preference). Here are my preliminary drawings. I don't think I will be changing them much. I apologize for the dimension clutter.
Any thoughts or suggestions would be appreciated. I will update this thread once the units have been built!
Attachments
Quick simulation of this baffle (approximating the U-baffle as a larger, but similarly shaped--flat baffle) in basta using the fostex specs and some PPA15 specs on the web (FWIW) suggests:
--This design will have a 2db hole centered on ~3K and a 3dB peak centered on ~1.5K.
--Extension--without EQ--will be to about 65hz at 3db down.
--Should be Smooth through crossover region (1db dip) --> though this does not include any floor bounce issues in the fostexes output just above XO
--About 6W power handling before the FF85 runs out of excursion at 200hz (PPA15 are only pushing 2.5mm at 20hz)
IMHO, The ripple at 1.5K, if not eq'ed, will be even worse if the rear reflection off the back wall is undamped.
The crossover point appears poorly chosen regarding maximum volume output, a single PPA15 per side would suffice, distortion at in the midrange 300hz-1K will be bad at volumes where the woofers are limping along.
325hz on the fostex and 200 on the woofers doubles the power-handling of the Fostex, but changes un-equed extension to ~75hz. Should be able to eq, reasonably flat to about 30hz if you use a 150w amp on the bottom and 15w on the top.
Sean
--This design will have a 2db hole centered on ~3K and a 3dB peak centered on ~1.5K.
--Extension--without EQ--will be to about 65hz at 3db down.
--Should be Smooth through crossover region (1db dip) --> though this does not include any floor bounce issues in the fostexes output just above XO
--About 6W power handling before the FF85 runs out of excursion at 200hz (PPA15 are only pushing 2.5mm at 20hz)
IMHO, The ripple at 1.5K, if not eq'ed, will be even worse if the rear reflection off the back wall is undamped.
The crossover point appears poorly chosen regarding maximum volume output, a single PPA15 per side would suffice, distortion at in the midrange 300hz-1K will be bad at volumes where the woofers are limping along.
325hz on the fostex and 200 on the woofers doubles the power-handling of the Fostex, but changes un-equed extension to ~75hz. Should be able to eq, reasonably flat to about 30hz if you use a 150w amp on the bottom and 15w on the top.
Sean
seanzozo said:
An interesting intellectual excercise, but the fostex data isn't that useful ... even for stock drivers -- these FF85KeN aren't stock. My sims show the uFonken show it xMax limited to 1w at 100 Hz... but playing unaided in my big room, i'm pretty sure it has a bit more than that in it. Early measures while i was developing Tysen showed no major dips (in room flat from 25 to 16k +/- 2 dB -- the simplier XO we ended up using is not as flat, but sounds better)
Personally, i think i'd be looking to have a bit more baffle to support the FF85keN as well as some assymmetry.
dave
seanzozo said:
Wow great response! Though I agree with planet10 in that the Fostex should have more headroom when playing music.
Reason I am doing two PPA15s per side is because, frankly I am a bit of a bass head and want to have the headroom.
Regarding te peak at 1.5K, what would you suggest I do? Add damping directly behind the baffle or a diffuser of some sort behind the driver on the back wall?
I will follow your advice on the crossover point though. I have the option to adjust the crossover point all the way up to 400hz for both the woofer and FR, so I can play around with it a lot.
As far as EQ, I will probably only use minimal EQ at the very lowest octaves.
To Dave, what do you suggest on baffle width? I don't want to go too wide, so I am thinking 8 inches wide and change the baffle geometry.
dave,
I expect that in a sealed box or a vented cabinet (as opposed to a baffle that provides a nulling effect at the excursion peak) the FF85 would have slightly greater headroom. Additionally, my point was not that a 6W simulated power-handling was too low in the abstract, but that mating that with 2 15-inch speakers was possibly overkill (unless you want to try EQ them flat below 30hz, then you literally cannot get enough woofer excursion
).
I think (agree) that either a greater effective-XO frequency or a wider baffle, would be reasonable. On the MJK OB baffle, adjusting the XO higher to meet the same XO freq, excursion is halved giving 6db additional mechanical headroom, while requiring essentially no EQ.
A wider asymmetric baffle will make the off-axis a little less organized and (0not shockingly) asymmetrical. But, in the narrower baffle the ~1.5K peak moves as you rotate off-axis (at 45 degrees it is centered at 1K) making the EQ less effective
Sean
I expect that in a sealed box or a vented cabinet (as opposed to a baffle that provides a nulling effect at the excursion peak) the FF85 would have slightly greater headroom. Additionally, my point was not that a 6W simulated power-handling was too low in the abstract, but that mating that with 2 15-inch speakers was possibly overkill (unless you want to try EQ them flat below 30hz, then you literally cannot get enough woofer excursion
).I think (agree) that either a greater effective-XO frequency or a wider baffle, would be reasonable. On the MJK OB baffle, adjusting the XO higher to meet the same XO freq, excursion is halved giving 6db additional mechanical headroom, while requiring essentially no EQ.
A wider asymmetric baffle will make the off-axis a little less organized and (0not shockingly) asymmetrical. But, in the narrower baffle the ~1.5K peak moves as you rotate off-axis (at 45 degrees it is centered at 1K) making the EQ less effective
Sean
I changed the dimensions a bit. I played with edge, roughly replicating the listening distance and width to get a ball park figure.
I increased the top edge to 8 inches from 4, and I removed the extreme curvature. I also moved the driver 1 inch up to be slightly offset from the center. The result is a response that closely resembles a rectangular baffle 8 inches with the same height. See attached image.
I still may add some curvature, but it will be very minimal compared to the original rendition. Again, I will raise the crossover point to 300hz+ for the FR.
Any more thoughts? I am usually not the one to spend too much time in the developing stages. I like to build So unless there is something vital I am missing other than the above mentioned, looks like the design is mostly finalized.
I increased the top edge to 8 inches from 4, and I removed the extreme curvature. I also moved the driver 1 inch up to be slightly offset from the center. The result is a response that closely resembles a rectangular baffle 8 inches with the same height. See attached image.
I still may add some curvature, but it will be very minimal compared to the original rendition. Again, I will raise the crossover point to 300hz+ for the FR.
Any more thoughts? I am usually not the one to spend too much time in the developing stages. I like to build
So unless there is something vital I am missing other than the above mentioned, looks like the design is mostly finalized.Attachments
I've been eyeing those Pyles for helper woofer duty with my Betsys. Ya' can't beat the price! I'll be really curious to see how their LF response turns out in that narrow baffle. MJK has shown the usefulness of the Alpha 15's considerably higher Q when designing OB's, but they ain't as cheap. I'd also be curious to hear about the Pyle's build quality from anybody who has laid hands on one.
pj
www.wildburroaudio.com
pj
www.wildburroaudio.com
I still think you need to offset the driver to the left or right and the baffle may not be wide enough, I'll sim it tonight with daves new numbers (they don't change much, when there are no XO components).
--threadjack alert--
Those numbers model well in the MJK-OB, with slightly modified XO components - I think I might use a pair in baffles I am going to do at some point for my mother-in-law.
Dave, how do they compare to the FX-120 in character, my mother-in-law heard my baffles and wants something similar?
Sean
--threadjack alert--
Those numbers model well in the MJK-OB, with slightly modified XO components - I think I might use a pair in baffles I am going to do at some point for my mother-in-law.
Dave, how do they compare to the FX-120 in character, my mother-in-law heard my baffles and wants something similar?
Sean
seanzozo said:
Sean,
Did you generate XO values? Share?
I have all the bits but the XOs for just this project.
Our FX120eNs haven't made it into a box yet. The treated FF85 are better than stock FX120 from as low as they go up (ie 300 Hz+) especially in downward dynamic range and detail. The 120 obviously go much lower, but the FF85KeN (compared to F120AeN) is more exciting, althou bat ears might find the rising top above 17k a bit annoying.
dave
Okseanzozo said:
I can definitely go wider on the baffle. I guess up to 24 inches at the widest point (currently its 16 as shown for the woofers)I am still open to suggestions, so keep them coming. My simulation skills are lacking. I really appreciate it
I also forgot to mention the power I have available. 50 watts per channel for the full range drivers and (for now) 75 watts for each pair of 15s.
With regard to fast-ones baffle. I suggest this as a starting point: a 20 inch x 42 inch baffle with 4 inch wings on the back.
The fostex sims best about 4 inches from the top and about 8 inches from the side
-- there is a very hi-Q 2db notch in the sim at 3K but otherwise it is +/- 1db from 100hz to 20K (with perfect drivers)
-- the bass rolls off at -3db at ~75hz
Why did I choose this size? In part it is working off of a known good, the MJKOB baffle. Basically the 1.5K peak in your original design has been pushed down (by widening the baffle) and spread out (by offsetting the driver) to the 400-800hz area so that it can be EQed-out by choosing an electronic XO point that is slightly higher than the intended acoustical XO.
Suggested starting points for the XO are 200hz and 425hz - acoustic XO should be around 320hz.
Caveats -
1) This sim does not include the floor or back wall. This means that real-world extension may be different (one kludge I used, mirroring the baffle suggests 3db extension might be up to 15hz lower than the basic sim, but in reality this will tend to be offset in part by the lack of a back wall in the sim). It also means that any nulling due to floor reflections is not included.
2) The parameters for the woofers are unverified.
3) The sim also approximated the wings as flat baffle extensions for the woofer, this probably understates their effect somewhat.
In all likelihood this baffle will work, being square it should be pretty easy to throw together cheaply and quickly. Bass extension will be moderate.
With regard to an FF85K/Alpha15 passive OB based on the MJKOB 20 inch x 38 inch baffle with driver placement as per his PDF
A first pass of XO components:
24uF and 3mh on the FF85K
10mH and 100uF on the Alpha 15
Use of the original parts (24uF+4mh and 9mH+68uF) results in a simulated 2db shelf up in the XO region (200-5550hz) because in the sim the FF85K tracks tighter phase with the woofer than the FE and FX series drivers. Acoustic XO looks to be about 340hz, a slightly lower XO than in my own FX120/Alpha15 baffle.
The fostex sims best about 4 inches from the top and about 8 inches from the side
-- there is a very hi-Q 2db notch in the sim at 3K but otherwise it is +/- 1db from 100hz to 20K (with perfect drivers)
-- the bass rolls off at -3db at ~75hz
Why did I choose this size? In part it is working off of a known good, the MJKOB baffle. Basically the 1.5K peak in your original design has been pushed down (by widening the baffle) and spread out (by offsetting the driver) to the 400-800hz area so that it can be EQed-out by choosing an electronic XO point that is slightly higher than the intended acoustical XO.
Suggested starting points for the XO are 200hz and 425hz - acoustic XO should be around 320hz.
Caveats -
1) This sim does not include the floor or back wall. This means that real-world extension may be different (one kludge I used, mirroring the baffle suggests 3db extension might be up to 15hz lower than the basic sim, but in reality this will tend to be offset in part by the lack of a back wall in the sim). It also means that any nulling due to floor reflections is not included.
2) The parameters for the woofers are unverified.
3) The sim also approximated the wings as flat baffle extensions for the woofer, this probably understates their effect somewhat.
In all likelihood this baffle will work, being square it should be pretty easy to throw together cheaply and quickly. Bass extension will be moderate.
With regard to an FF85K/Alpha15 passive OB based on the MJKOB 20 inch x 38 inch baffle with driver placement as per his PDF
A first pass of XO components:
24uF and 3mh on the FF85K
10mH and 100uF on the Alpha 15
Use of the original parts (24uF+4mh and 9mH+68uF) results in a simulated 2db shelf up in the XO region (200-5550hz) because in the sim the FF85K tracks tighter phase with the woofer than the FE and FX series drivers. Acoustic XO looks to be about 340hz, a slightly lower XO than in my own FX120/Alpha15 baffle.
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.